Why explain? Why should it be a major objective of scholars to identify relationships that are explanatory? One reason, usually considered self-justifying, is to satisfy curiosity. Another reason, and probably the more common one, is to provide a basis for prediction. And prediction in turn is prized as a basis for making choices.
In discussing explanation-and in other connections, too--we have mentioned choices. Scholarship reflects choices. We have argued that the choices should be such as to make the scholarship of a desirable sort, i.e., significant, or contributory (at least potentially) to rationality in decision making. Assuming agreement on this, suppose that the scholar faces a choice. He has two alternative lines of action before him. How does he go about deciding between them? Obviously, he makes one or more predictions. He may predict that one alternative is surer to lead to significant results than the other, or that one will lead to more significant results than the other. Having made his prediction(s), he makes his choice. The choice turns on the prediction(s), considered in terms of the goal being pursued.
These statements can be made more general. Whenever anyone makes a rational choice, he is predicting. Those who chose rationally to vote for Eisenhower rather than for Stevenson in 1956 were presumably predicting that in some sense Eisenhower would make a more desirable President. A Congressman who votes for a bill